5 citations · 10 across the 18 of their papers we have counts for
11 papers · 1 filter
GRPO Training Dynamics for Small Language Models
Rajat Ghosh, Vaishnavi Bhargava, Henry Wong +2
Group Relative Policy Optimization (GRPO) has emerged as a memory-efficient reinforcement fine-tuning (RFT) technique for reasoning-intensive tasks. How- ever, GRPO training dynami…
Reduced-Order Models: The Mother of World Models
Rajat Ghosh
World models -- compressed latent representations of an environment that support action-conditioned prediction and planning -- are typically presented as a product of modern self-s…
Predictable GRPO: A Closed-Form Model of Training Dynamics
Rajat Ghosh, Datta Nimmaturi, Aryan Singhal +4
We develop a first-principles reduced-order model of these dynamics. Under a single mean-field assumption that summarizes the policy by its expected reward, we reduce the GRPO upda…
Croissant Baker: Metadata Generation for Discoverable, Governable, and Reusable ML Datasets
Rafi Al Attrach, Rajna Fani, Sebastian Lobentanzer +17
Croissant has emerged as the metadata standard for machine learning datasets, providing a structured, JSON-LD-based format that makes dataset discovery, automated ingestion, and re…
Action Shapley: A Training Data Selection Metric for World Model in Reinforcement Learning
Rajat Ghosh, Debojyoti Dutta
Numerous offline and model-based reinforcement learning systems incorporate world models to emulate the inherent environments. A world model is particularly important in scenarios…
Go-UT-Bench: A Fine-Tuning Dataset for LLM-Based Unit Test Generation in Go
Yashshi Pipalani, Hritik Raj, Rajat Ghosh +2
Training data imbalance poses a major challenge for code LLMs. Most available data heavily over represents raw opensource code while underrepresenting broader software engineering…
A Multi-Agent Framework for Stateful Inference-Time Search
Arshika Lalan, Rajat Ghosh, Aditya Kolsur +1
Recent work explores agentic inference-time techniques to perform structured, multi-step reasoning. However, stateless inference often struggles on multi-step tasks due to the abse…
BAR Conjecture: the Feasibility of Inference Budget-Constrained LLM Services with Authenticity and Reasoning
Jinan Zhou, Rajat Ghosh, Vaishnavi Bhargava +2
When designing LLM services, practitioners care about three key properties: inference-time budget, factual authenticity, and reasoning capacity. However, our analysis shows that no…
Predictive Scaling Laws for Efficient GRPO Training of Large Reasoning Models
Datta Nimmaturi, Vaishnavi Bhargava, Rajat Ghosh +2
Fine-tuning large language models (LLMs) for reasoning tasks using reinforcement learning methods like Group Relative Policy Optimization (GRPO) is computationally expensive. To ad…
Efficient Alignment of Large Language Models via Data Sampling
Amrit Khera, Rajat Ghosh, Debojyoti Dutta
LLM alignment ensures that large language models behave safely and effectively by aligning their outputs with human values, goals, and intentions. Aligning LLMs employ huge amounts…
Data-Driven Evaluation of Training Action Space for Reinforcement Learning
Rajat Ghosh, Debojyoti Dutta
Training action space selection for reinforcement learning (RL) is conflict-prone due to complex state-action relationships. To address this challenge, this paper proposes a Shaple…